Dissertations

This dissertation applies machine learning in physics, focusing on theory and practical implementation. In molecular physics, the absence of ionization cross section (ICS) data for electron collisions with specific molecules poses a significant challenge. ICS measurements are crucial for understanding electron-molecule interactions, pivotal in atmospheric science, radiation chemistry, and material development. Neural networks are used to predict ICS values for molecules within families that share common constituent atoms (e.g., SF_x for x = 1, 2, ..., 6.

This report explores Quantum Field Theory in Rindler spacetime, beginning with Special Relativity and Penrose's spacetime representation. We analyze Momentarily Comoving Reference Frames (MCRFs) and proper acceleration, focusing on the Rindler observer with constant acceleration. Quantizing a scalar field in Minkowski and Rindler spaces reveals two vacuum states, where the Rindler observer perceives the Minkowski vacuum as filled with particles—the Unruh effect. The Unruh temperature is measured via Bogolyubov transformations between vacuum states.

We investigate the cosmic microwave background (CMB), studying tensor algebra, calculus, general relativity basics, and cosmology. Using Scilab's symbolic package, we develop tools to compute key tensors like Christoffel symbols and derive equations for the Friedmann-Lemaître-Robertson-Walker metric. We also model CMBR with a circular patch divided into 2, 4, and 8 sections representing dipole, quadrupole, and octupole temperature distributions, aiming to plot their power spectrum using Scilab.

The study delves deep into quantum mechanics, emphasizing concepts like non-locality and entanglement. It covers key developments such as uncertainty, non-locality, entanglement, and hidden variables, along with topics like the EPR paradox, Bell's Inequality, and quantum experiments by researchers like Aspect et al.

This report computationally models relativistic and non-relativistic effects on single electron systems using the shooting method alongside the fourth-order Runge-Kutta method. We approximate eigenenergies by plotting boundary functions and refine these using the Newton-Raphson Method. Our findings include precise calculations of transition energies, binding energies, and energy eigenvalues and eigenfunctions, achieving an accuracy within 10^{-2} eV for transition energies.